That is indeed a very strange problem. I'll have a look tomorrow to see if I can reproduce this bug — it's certainly not something I have seen or had reported before.
Would you mind testing with a default .zshrc or sending me your .zshrc in case it is related to a Ash setting you have turned on that I don't?
Many thanks,
Paul
Thank you for investigating this further. I'll have a look through your config though I can't think what it could be.
The transport not connected error I get if the process hosting the VFS errors. What I tend to do is run this process manually instead of using 'tmsu mount'. If you wouldn't mind trying this I would be very interested in the error shown. The syntax is as follows:
>tmsu vfs path/to/db path/to/mount point
The process runs foreground and should write a stack trace when it dies, assuming that it is dying.
Paul
Tried reproducing the error with an empty .zshrc, everything works just fine. So the problem probably is due to some weird config setting. My zsh config is on github, see https://github.com/dadrc/zsh-cfg
The two functions involved above look as follows, in case there is something obviously wrong about them.
ls: aliased to ls -b -CF --color=auto
cd () {
if (( ${#argv} == 1 )) && [[ -f ${1} ]]
then
[[ ! -e ${1:h} ]] && return 1
print "Correcting ${1} to ${1:h}"
builtin cd ${1:h}
else
builtin cd "$@"
fi
}
I will try to track down the offending setting and report here if I'm successful.
Thanks for the hint,
Phil
PS: I didn't get your answer mail from the mailing list although I subscribed, so this answer comes to you straight from Google Groups, I hope this doesn't screw up your thread view.
On Saturday, 21 July 2012 23:19:58 UTC+2, Paul Ruane wrote:
That is indeed a very strange problem. I'll have a look tomorrow to see if I can reproduce this bug — it's certainly not something I have seen or had reported before.
Would you mind testing with a default .zshrc or sending me your .zshrc in case it is related to a Ash setting you have turned on that I don't?
Many thanks,
Paul
It seems then that something in your .zshrc is checking whether there is a '.git' directory each time you run 'ls' (for whatever reason). This perfectly reasonable behaviour, though for your own peace of mind you may want to track it down.
Because the TMSU VFS is only expecting operations on valid tags, this is causing the error. Ultimately it's the fault of TMSU (and therefore me) as what it should be doing is reporting that the directory does not exist rather than dying.
I'll push out a fix for this in the next few days. Apologies for the hassle it has caused you and thanks for helping me track it down.
Paul